Bun 1.4 Rust rewrite is not looking good

Bun, a project that gained traction since its initial release in 2022, has recently faced challenges with its 1.4 Rust rewrite. This rewrite, which started in the past three months, has been met with disappointment and frustration from the community.

Once a symbol of progress and feature implementation, the project has now become associated with continuous false promises. After the Rust rewrite, the frequency of positive tweets about upcoming releases has dwindled, replaced by vague communication and unfulfilled expectations.

The user reaction to these false promises has been predictable: a shift in developers' allegiance. One user, @jarredsumner, announced they were switching to Go, expressing frustration with the repeated false promises of future releases.

The Bun 1.4 rewrite is, in fact, a significant bet on AI. With 15.8k commits from robobun, 1.6k from autofix-ci[bot], and 790 from Jarred, the project has seen an unprecedented amount of activity. However, the majority of PRs now originate from Claude prompting itself, rather than human contributors.

The project currently boasts over 5,000 open pull requests, a staggering number that warrants caution. GitHub recommends limiting open PRs to under 1,000 per branch to avoid mergeability checks timing out.

Critics argue that the rewrite has not delivered on its promise of improved memory safety. Instead, the rewrite appears to push an Anthropic agenda, questioning if AI agents can effectively manage a production codebase with minimal human oversight.

The rewrite, critics suggest, may have been driven by early decisions to use Rust, rather than the inherent issues of Zig that initially prompted the rewrite. A rewrite into Zig, they argue, might have been a more sensible alternative. It remains to be seen if Bun can recover from this long tail of issues stemming from the failed rewrite.